Differences in different reconstruction algorithms for coronary CTA demonstrating pericoronary adipose tissue...

Area of Science:

• Cardiovascular Imaging
• Medical Physics
• Radiology

Background:

• The Fat Attenuation Index (FAI) is a sensitive biomarker for coronary inflammation, measurable via Coronary Computed Tomography Angiography (CCTA).
• Different image reconstruction algorithms used in CCTA may influence FAI measurements.

Purpose of the Study:

• To evaluate differences in FAI measurements on CCTA using three reconstruction algorithms: Deep Learning Image Reconstruction (DLIR-H), adaptive statistical iterative reconstruction-Veo (ASiR-V50%), and Filtered Back Projection (FBP).

Main Methods:

• Patients were categorized based on coronary plaque presence (normal, no plaque, non-calcified, mixed, calcified).
• FAI, pericardial adipose tissue attenuation, and image noise were measured for each algorithm (DLIR-H, ASiR-V50%, FBP) across coronary arteries.
• Statistical analysis was performed to compare measurements between algorithms and plaque groups.

Main Results:

• Pericardial adipose tissue attenuation showed no significant difference between algorithms.
• FAI values were highest with DLIR-H, intermediate with ASiR-V50%, and lowest with FBP, with all differences being statistically significant.
• DLIR-H exhibited superior image noise reduction compared to FBP.

Conclusions:

• The choice of reconstruction algorithm significantly impacts FAI measurements in CCTA.
• Standardized reconstruction protocols are essential for multicenter and longitudinal studies to ensure reproducible and comparable FAI results.